Microarrays and DNA Sequencing Improve Prenatal Testing

Ricki Lewis, PhD

December 05, 2012

Prenatal diagnosis is becoming more precise, according to 3 reports published in the December 5 issue of the New England Journal of Medicine. Chromosomal microarray (CMA) analysis is augmenting standard karyotyping of fetuses and stillbirths, as genome sequencing reveals disrupted disease-causing genes at the breakpoints of balanced rearrangements.

"These results highlight the power and complexity, and some of the pitfalls, of using new genomic technology in clinical practice," Lorraine Dugoff, MD, from the Department of Obstetrics and Gynecology, University of Pennsylvania Perelman School of Medicine, Philadelphia, writes in an editorial accompanying the articles.

CMAs use DNA probes to detect deletions and repeats. These copy number variants (CNVs) are from 1000 to 5 million bases long, lying beneath the highest resolution of the stained bands and fluorescent labels of conventional cytogenetics.

A major advantage of CMA is that samples need not be cultured, which can introduce artifacts, and can be degraded. However, CMA is blind to chromosomal rearrangements that do not alter copy number, such as balanced translocations and inversions, as well as to polyploidy.

Since 2005, CMA has been the first-tier diagnostic test for autism spectrum disorder, developmental delay, multiple congenital structural anomalies, and intellectual disability in pediatrics. Prenatal applications of the test began in 2006.

CMA finds a genetic cause in 15% more cases than does karyotyping.

Prenatal Testing

In 2009, the American College of Obstetrics recommended CMA if ultrasound is abnormal but karyotype is normal. CMA might even become a standard part of invasive testing (amniocentesis and chorionic villus sampling).

"It's hard to justify offering a routine karyotype alone when CMA can provide significant additional information with no additional procedural risk," Andy Faucett, MS, CGC, director of policy and education at the Genomic Medicine Institute at Geisinger Health System in Pennsylvania, told Medscape Medical News. Dr. Faucett was not involved in any of the newly reported studies.

In the first article, Ronald Wapner, MD, from Columbia University, New York City, and colleagues compared CMA to karyotyping for 4406 samples from women attending prenatal diagnosis centers for advanced maternal age, trisomy detected in maternal serum, or abnormal ultrasound. They obtained CMA results for 4340 (98.8%) of the samples. Comparison with parental DNA identified de novo CNVs.

The researchers classified CNVs as pathogenic, likely benign, or "variants of uncertain clinical significance." Patients were told about pathogenic variants, but not about likely benign ones. An independent clinical advisory group identified the uncertain variants that have "potential clinical significance," and patients were informed.

CMA detected all aneuploidies seen on karyotypes, in addition to smaller CNVs and deletions. It identified clinically relevant deletions or duplications in 6.0% of fetuses with normal karyotypes but abnormal ultrasound, and in 1.7% of fetuses with normal karyotypes and older mothers or a trisomy.

Overall, CMA identified 1399 prenatal samples with CNVs, and of those samples, 1234 (88.2%) were benign and 35 (0.9%) were pathogenic. Of the remaining 130 samples, 36 were likely benign and 94 were of uncertain significance. The committee identified 61 (64.9%) of 94 of those samples as alarming enough to tell the patient.

Reporting seemingly abnormal CNVs without a clear diagnosis is not ideal. "Some patients don't want uncertainty and aren't willing to get additional incremental information, whereas others say that additional information is way more important than the possibility of uncertainty. Patient autonomy is important," Dr. Wapner told Medscape Medical News.

However, he also points out that CMA results can be actionable. "If you have a CNV that predisposes a child to learning disabilities, early intervention may be helpful." The researchers will follow up the children with CNVs to inform future diagnoses.


In the second article, Uma M. Reddy, MD, MPH, from the Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, Maryland, and colleagues used CMA to identify genetic causes of stillbirths.

Dr. Reddy and colleagues evaluated 532 samples with postmortem exams, karyotyping, and CMA for CNVs smaller than 500 kb. They excluded CNVs in databases of non–disease associated variants and classified the remainder as probably benign, clinical significance unknown, or pathogenic.

The researchers obtained microarray results for 465 (87.4%) of the 532 stillbirths compared with 375 (70.5%) of the 532 stillbirths for karyotypes (P < .001), likely because microarrays work on degraded tissue. The CMA results included many benign variants and also more aneuploids and pathogenic CNVs (8.3% vs 5.8%; P = .007) than did karyotyping.

For the 157 stillbirths without karyotypes, CMA had a 79.6% success rate. Of those stillbirths, 73.9% were normal or probably benign variants and 5.7% were abnormal.

"For couples with a pregnancy loss, knowing why is extremely important. CMA testing will greatly increase the chance that couples will have an answer and will allow them to plan for future pregnancies," Faucett told Medscape Medical News. Testing parents will be necessary to distinguish de novo mutations from inherited ones, which the study did not do.

When Are Karyotypes Needed?

The third article, from Michael E. Talkowski, PhD, from the Center for Human Genetic Research, Massachusetts General Hospital, Boston; the Program in Medical and Population Genetics, Broad Institute, Cambridge; and the Department of Neurology, Harvard Medical School, Boston, Massachusetts, and colleagues, supports continued karyotyping. The authors report a case in which genome sequencing identified a mutation in a gene (CHD7) that caused CHARGE syndrome in a fetus with normal CMA results but an abnormal ultrasound and a balanced translocation on karyotype.

"Prenatal microarrays won't detect cases like this. And if you don't do a karyotype, you won't know you have a balanced rearrangement. The next pregnancy could be unbalanced," lead author Cynthia Morton, PhD, director of cytogenetics at Brigham and Women's Hospital, Boston, told Medscape Medical News.

Even as chromosomal microarrays and classic karyotyping reveal ever more, obsolescence looms for both. "We are in a transitional period. It's only a window of time until whole-genome sequencing picks up deletions, duplications, and balanced rearrangements," Dr. Morton concluded.

Authors of all 3 studies disclosed a variety of financial conflicts including relationships with Agilent Technologies, Affymetrix, Signature Genomics, Genzyme, Shire, CombiMatrix, NimbleGen, Celula, and Novartis. Full conflict-of-interest information is available on the journal's Web site. Dr. Dugoff received a research award from PerkinElmer for a different study. Faucett has disclosed no relevant financial relationships.

N Engl J Med. 2012;367:2175-2184, 2185-2193, 2226-2232, 2249-2251.Wapner article, Reddy article, Talkowski, Editorial